The current paradigm for glioblastoma (GBM) therapy is based on the concept that each patient is treated with a protocol that is most likely to prolong life (temozolomide and radiation) in a majority of patients. The treatment may not be efficacious for any given individual even though it may work with most patients with a similar pathology. If the treatment fails, second-line treatments, usually experimental are administered, although by this time it may be too late for many GBM patients. This concept is obviously unacceptable in the current era wherein the technology for whole genome sequencing is a reality thus enabling personalized medicine using targeted agents based on genetic alterations of each patient is possible. Unfortunately this concept is not financially practical. Project 1 of the P01 will investigate if stratification of patients based on our current understanding of dominant GBM subtypes will lead to significantly improved outcomes. Specific Aim 1 will utilize a set of primary patient derived xenograft models each of which has been classified according to the molecular signatures of GBM using genetic profiling. The response of these human explants to the standard of care (SOC) therapy will be investigated to establish the relative sensitivities of each GBM subclass to SOC. Aim 1 will also investigate the genetic basis for the three subclasses of response to SOC observed in our recent clinical trial using imaging as a biomarker. Specific Aim 2 will extend these studies to targeted single agent therapeutic paradigms for each of the molecular subtypes using cell based assays and in vivo orthotopic primary mouse models. Since GBM is a heterogeneous disease targeted inhibition of a single agent may be compensated by activation of compensatory oncogenic pathways. Based on the above studies. Specific Aim 3 will design and validate combination therapies for each of the molecular subtypes. Based on our existing expertise in mouse models of cancer and the application of molecular imaging combined with proteomics and genomics to rationally design therapeutic paradigms for distinct molecular subclasses of GBM, Project 1 will provide proof of principle results that will allow stratification of patients based on their molecular signature. This we believe will ultimately result in greatly improved outcomes for this disease. The ultimate goal of this project is to impact the patient outcomes through an optimized/rationale design of clinical trials in Project 3 wherein imaging biomarkers are paired with the most efficacious combination therapies for each of the GBM subtypes during clinical translation.